Reactive magnetron sputtering of tungsten target
in krypton/trimethylboron atmosphere
Martin Magnuson, Lina Tengdelius, Fredrik Eriksson, Mattias Samuelsson, Esteban Broitman, Grzegorz Greczynski, Lars Hultman and Hans Högberg
The self-archived postprint version of this journal article is available at Linköping University Institutional Repository (DiVA):
http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-160243
N.B.: When citing this work, cite the original publication.
Magnuson, M., Tengdelius, L., Eriksson, F., Samuelsson, M., Broitman, E., Greczynski, G., Hultman, L., Högberg, H., (2019), Reactive magnetron sputtering of tungsten target in krypton/trimethylboron atmosphere, Thin Solid Films, 688, 137384. https://doi.org/10.1016/j.tsf.2019.06.034
Original publication available at:
https://doi.org/10.1016/j.tsf.2019.06.034 Copyright: Elsevier
1
Reactive magnetron sputtering of tungsten target in
krypton/trimethylboron atmosphere
Martin Magnuson, Lina Tengdelius, Fredrik Eriksson, Mattias Samuelsson, Esteban Broitman,† Grzegorz Greczynski, Lars Hultman, and Hans Högberg
Department of Physics, Chemistry, and Biology (IFM) Linköping University, SE-581 83 Linköping, Sweden
†Present address: SKF Research and Development Center, 3439 MT Nieuwegein,
Netherlands
2019-06-19
*Corresponding author: martin.magnuson@liu.se
Abstract
W-B-C films were deposited on Si(100) substrates held at elevated temperature by reactive sputtering from a W target in Kr/trimethylboron (TMB) plasmas. Quantitative analysis by X-ray photoelectron spectroscopy (XPS) shows that the films are W-rich between ~ 73 and ~ 93 at.% W. The highest metal content is detected in the film deposited with 1 sccm TMB. The C and B concentrations increase with increasing TMB flow to a maximum of ~18 and ~7 at.%, respectively, while the O content remains nearly constant at 2-3 at.%. Chemical bonding structure analysis performed after samples sputter-cleaning reveals C-W and B-W bonding and no detectable W-O bonds. During film growth with 5 sccm TMB and 500 oC or with 10 sccm
TMB and 300-600 oC thin film X-ray diffraction shows the formation of cubic 100-oriented
WC1-x with a possible solid solution of B. Lower flows and lower growth temperatures favor
growth of W and W2C, respectively. Depositions at 700 and 800 oC result in the formation of
WSi2 due to a reaction with the substrate. At 900 oC, XPS analysis shows ~96 at.% Si in the
film due to Si interdiffusion. Scanning electron microscopy images reveal a fine-grained microstructure for the deposited WC1-x films. Nanoindentation gives hardness values in the
range from ~23 to ~31 GPa and reduced elastic moduli between ~220 and 280 GPa in the films deposited at temperatures lower than 600 oC. At higher growth temperatures the hardness
decreases by a factor of 3 to 4 following the formation of WSi2 at 700-800 oC and Si-rich surface
at 900 oC.
Keywords: W-B-C films, reactive magnetron sputtering, trimethylboron, nanoindentation, X-ray photoelectron spectroscopy, thin film X-X-ray diffraction, Scanning Electron Microscope
2
1. Introduction
Tungsten carbide and the d-WC phase [1] with a hexagonal stacking sequence (Strukturbericht designation Bh) is a technologically important material best illustrated from its use on cemented
carbide tools for metal machining. Favorable to applications, the d-WC property envelope includes: a melting point of 2870 oC, Vickers hardness (HV) of 22 GPa, Young’s modulus of
elasticity 620-720 GPa, and good oxidation resistance [2]. If d-WC could be grown as a thin film material the number of applications would increase substantially by exploiting for instance the carbide’s low resistivity of 17-22 µW.cm [2]. A low resistivity in combination with the
previously defined properties is thus favorable in high-temperature Schottky contacts [3]. However, the literature is clear on the difficulties in depositing d-WC by chemical vapor deposition (CVD) or physical vapor deposition. Presently, CVD seems to be the most promising technique for growth of films containing the phase [4] [5] [6] [7] [8]. The high temperatures typical for these studies with 1200 oC [4], 950 oC [6], and 900 oC [7] [8] as well as the restriction
to W(110) substrates [5] limits the number of applications and calls for alternative deposition techniques. In addition, competing phases are frequently present in the deposited films seen from the tungsten rich carbide W2C and the cubic and substoichiometric g-WC with NaCl-type
structure (Strukturbericht designation B1) [1], henceforth referred to as WC1-x. The properties
for the metastable WC1-x are, however, not as favorable as those for d-WC seen from carbon
vacancies with x~0.6 [1]. In carbides with B1 structure such as WC1-x carbon vacancies will
result in lower hardness [2] [9].
Magnetron sputtering offers the possibility for low-temperature growth and the literature shows that WC films have been deposited from WC targets, [10] [11] [12] [13], W and C sources [14] [14] [15] and of particular interest for this study by reactive sputtering. The most applied gaseous reactants have been C2H2 (acetylene), [11] [16] [17] [18] [19] [20] and CH4 (methane)
[21] [22], but with report from reactive processing from other hydrocarbons such as ethene (C2H4) [12] and benzene (C6H6) [23]. As experienced from CVD, there are difficulties in
depositing the d-WC phase with only a few reports on reactive sputtering of phase-pure films, using alloying with N2 [24] or growth on carbon substrates at temperatures > 900 oC [12].
Thus, the prospect of depositing single-phase d-WC films by sputtering seems difficult. For this task, alloying by a third element such as B constitutes a viable route to improve the properties of WC1-x films by filling C vacancies with B atoms. Liu et al. have reported reactive sputtering
from a WB2 target in C2H2 containing plasmas for deposition of W-B-C films [25]. The study
shows that the composition of the films can be varied at low flows of C2H2, but with deposition
of almost pure carbon films when the WB2 target is poisoned by C2H2. In two publications,
Alishahi et al. [26] and Debnárová et al. [27] investigated growth of W-B-C films from magnetron sputtering of W, and B4C targets, in combination with pulsed sputtering of a C target.
The applied experimental set-up with three separate sources allows for growth of films in a with a wide range of compositions and properties as reported by the authors but is less practical for industrial applications.
An alternative and more industrially compatible route is to apply boron as a gaseous reactant and then preferably as a single precursor containing both carbon and boron. In a seminal study, Lewis et al. [28] compared trialkylboron triethylboron [TEB, B(C2H5)3], trimethylboron (TMB,
B(CH3)3) and tributylboron [TBB, B(C4H9)3] and suggested that TEB was suitable for
depositing boron carbon films by CVD. In addition, the authors have successfully applied TEB [29] [30] [31] for growth of epitaxial sp2-BN films by thermally activated CVD. The fact that
3
is a solid makes TMB the easiest trialkylboron to integrate in a sputtering process. In addition, TMB exhibits the highest boron-to-carbon of the organoboranes with 1:3 compared to 1:6 in TEB and 1:12 in TBB, which allows for studying boron-rich W-B-C films.
To investigate the possibility of integrating a common precursor from CVD in magnetron sputtering, we reactively sputter a W target with TMB as a single gaseous reactant for growth of W-B-C films. The flow of TMB is varied for deposition of W-rich films and with growth carried out without external heating (room temperature, RT) to 900 oC.
2. Experimental details
The W-B-C films were deposited in an ultra-high vacuum system (base pressure 2 x 10-6 Pa) on Si(100) substrates by reactive sputtering of a 3 inch circular W target in a Kr plasma (99.998%) using TMB gas with a purity of 99.99% from Voltaix Inc., High Springs, FL, USA as precursor. Krypton was applied as sputtering gas instead of Ar to reduce the probability of back-scattered neutrals that will introduce stresses in the films. All films were deposited for 5 min without substrate rotation, using a sputtering current of 900 mA and with the substrates held at floating potential. In our depositions the TMB gas was distributed close to the substrate table through a gas pipe mounted in the deposition chamber and with the Kr gas was distributed in a separate line at a flow of 82 sccm in all depositions. This set-up was chosen to minimize poisoning of the W target by TMB. The substrate was mounted ~5 cm from the end of the gas pipe and with the substrate positioned directly in line-of-sight above the magnetron at a distance of 7 cm. The substrate was heated by a heating stage positioned directly above the substrates. For additional information on the applied deposition system the reader is referred to ref. [32]. For films deposited with 10 sccm TMB in the plasma the deposition temperature was investigated from RT, 100, 200, 300, 400, 500, 600, 700, 800, and 900 °C. At 500 °C the TMB content in the plasma was varied from 1, 2.5, 5, 7.5, and 10 sccm. The addition of TMB to the plasma caused a slight increase of the total pressure from 0.53 Pa for sputtering without TMB in the plasma to 0.6 Pa at the highest applied flow of TMB as well as increased voltage on target as dependent on the amount of TMB in the plasma. The pronounced poisoning of the target seen from an increasing target voltage from ~450 V to ~600 V limited the applied flow of TMB to 10 sccm in this study. Prior to deposition, the substrates were ultrasonically degreased for 5 min in trichloroethylene, followed by 5 min in acetone, and finally 5 min in isopropanol. The chemical composition and the chemical bonding structure of the films was assessed by X-ray photoelectron spectroscopy (XPS). The instrument used was an AXIS Ultra DLD from Kratos Analytical, with monochromatic Al Kα radiation (hν = 1486.6 eV), operated at a base pressure of 1.5×10-7 Pa and with the X-ray anode at 225 W. The binding energy scale was
calibrated by setting the position of the Fermi edge of a sputter-cleaned Ag sample to 0.0 eV [33] resulting in the position of the Ag 3d5/2 core-level peak of 368.30 eV [34]. To remove
adsorbed contaminants following air exposure, the samples were sputter-cleaned for 180 s with 4 keV Ar-ions incident at an angle of 70° with respect to the surface normal. Casa XPS software (version 2.3.16) was used for quantification, with elemental sensitivity factors supplied by Kratos Analytical Ltd. The confidence level of XPS is typically around ±5 %. X-ray diffraction (XRD) θ/2θ scans were used to evaluate the structural properties of the films using a Philips PW 1820 Bragg-Brentano diffractometer, equipped with a Cu anode X-ray tube (Cu Kα, l=1.54
Å) operated at 40 kV and 40 mA. Additional XRD scans were performed in grazing incidence (GI) geometry with an incidence angle of a=2°, using a Panalytical Empyrean diffractometer in a parallel beam setup with a line focus Cu-Ka X-ray source operating at 45 kV and 40 mA.
4
The primary beam was conditioned using an X-ray mirror and a ½° divergence slit, and in the secondary beam path a 0.27° parallel plate collimator was used together with a PIXcel detector in 0D mode for data acquisition. XRD pole figure measurements were
performed using the same
Empyrean diffractometer although now with a point-focused copper anode source and an X-ray lens with 2x2 mm2 crossed-slits on the
incident beam side to reduce the effect of defocussing when tilting the sample. The detector position was fixed at specific diffraction angles, corresponding to diffraction from WC1-x {200}, {220}, and {222} family of planes, respectively. The pole figure
measurements were performed in 5°-steps with azimuthal rotation 0° ≤ f ≤ 360° and tilting 0° ≤ y ≤ 85° to determine the orientation distribution of the crystals.
Film morphologies and thicknesses were investigated using cross-sectional scanning electron microscopy (SEM). The instrument was a LEO 1550 Gemini SEM and the applied acceleration voltage was 10 kV. The nanomechanical properties of the films were measured by quasi-static displacement-controlled nanoindentation tests, using a Hysitron Triboindenter model TI950. Hardness (H) and reduced elastic modulus (Er), were measured with a Berkovich diamond
probe and calculated according to the method proposed by Oliver and Pharr [35]. The tip area function was calibrated using a fused silica sample. The thermal drift was compensated prior to each measurement, and no pile-up corrections were necessary. For each test, a total of 12 indents with a spacing of 10 µm were averaged to determine the mean value and standard deviations of H and Er.
The “rule of thumb” for nanoindentations, stating that the indenter should not penetrate more than 10% of the total film thickness, has been shown to fail in many materials (see [36] and references therein). In some of our W-B-C films, the maximum penetration depth hmax was
calculated using the graphical method explained in the standard ISO 14577 Part 4 [37]. This was validated by performing numerous indentations in one of the harder films at different penetration depths to conclude that the displacement-controlled nanoindentation experiments should be carried out in our samples at hmax = 70 nm indentation depth.
To be able to relate HV reported by others to our nanoindentation values given in GPa, an approximate conversion of HV values to GPa were made by multiplying the Vickers hardness values by 0.009807, i.e. converting kg/m2 to Pa. The exact conversion from a Vickers hardness
into the nanoindentation hardness values needs a geometrical factor correction which is ~0.927 for a perfect Berkovich diamond [36]. It should be noted that the relation is correct only for materials that deform fully plastically during the indentation. If the material has an elastic recovery, the relation is not valid because the Vickers indentation hardness is calculated using the deformed area after indentation, and in the nanoindentation method the hardness is defined by using the deformed area at maximum applied load [36]. To compare reported values on elastic modulus (E) from the literature with our measured Er values, we applied the formula;
1/Er=((1-n2)/E)+((1- ni2)/Ei)) from ref. [36] and assumed: a Poisson’s ratio for the sample
Figure 1: Elemental compositions for W-B-C films deposited at 500 oC and with TMB flows of 1, 2.5, 5, 7.5 and 10 sccm in (a) and for films deposited at RT, 100, 200, 300, 400, 500, and 600 oC and with 10 sccm TMB in the plasma in (b). 20 15 10 5 0 Content (at.%) 10 8 6 4 2 0 TMB flow (sccm) 20 15 10 5 0 Content (at.%) 600 400 200 0 Temperature (oC) 100 90 80 70 60 100 90 80 70 60 W C B O a) b)
5
n=0.25, a Poisson’s ratio for the indenter of n=0.07, and the elastic modulus of the indenter to be Ei=1140 GPa.
3. Results and Discussion
3.1 Composition and chemical bonding structure
Figure 1a shows the elemental
composition obtained from the
quantitative analysis of XPS spectra for W-B-C films deposited with TMB flows of 1, 2.5, 5, 7.5, and 10 sccm at 500 oC.
As observed, all films are metal-rich and where the W content decreases from ~ 93 at% to ~ 73 at% from with increasing TMB content in the plasma. As expected, the C and B contents in the films increase with increasing TMB flows from ~ 4 to ~ 18 at.% for C and ~ 0 to ~ 7 at.% for B. The C-to-B ratio is 3:1 for the film deposited with 10 sccm TMB i.e., the same as in the applied TMB precursor, while films deposited with 5 and 7.5 sccm TMB show a more boron rich composition with C/B~2.5, but close to a 3:1 ratio. The film deposited with 1 sccm TMB is carbon rich as evident from C/B~5, which is probably due to uncertainties in determining the B content given the low intensity of the B 1s peak in XPS. The O content is nearly constant at 2-3 at.%.
Note that all layers were exposed to air before XPS measurements which resulted in a few nm thick native oxides at the surface. Prior to XPS analyses all samples were cleaned with Ar ion beam. Hence, the oxygen detected in the films can have two sources: (1) redeposited atoms following the sputter-etch, and (2) implanted atoms resulting from a forward momentum transfer during Ar sputter-cleaning. Figure 1b shows that the W, C, B, and O content remain nearly constant for films deposited at RT, 100, 200, 300, 400, 500, and 600 oC and with 10
sccm TMB in the plasma. The W content is between 75 to 76 at.% for films deposited at RT to 300 oC and between 72 to 73 at.% for films deposited at 400, 500 and 600 oC. At higher
deposition temperatures, Si is detected in the films (not shown) suggesting a reaction with the Si(100) substrate, which is further discussed with the XRD results. Carbon shows an opposite trend compared to W with about 15 at.% below 300 oC and 17-18 at.% above 300 oC, while no
clear trend is found for B that varies between 6-8 at.%. We suggest that the difference in W and C content in the films is due to a more efficient dissociation of the TMB precursor at higher temperatures, resulting in a slightly carbon-rich composition in the temperature range from 300 to 600 oC. Similar as for the films deposited with different flows of TMB in the plasma the
oxygen content is constant at 2-3 at.%.
Figure 2a and 2b show W 4f and C 1s core-level XPS and Fig. 3a and 3b show B 1s and O 1s for W-B-C films deposited with TMB flows of 1, 2.5, 5, 7.5, and 10 sccm at 500 oC. The W 4f7/2 and 4f5/2 peaks in Fig. 2a, indicated by the vertical dotted lines, are located at 31.4 eV and
33.6 eV, respectively, with a spin-orbit splitting of 2.2 eV. These values are identical to those reported in the literature for metallic W [38], thus supporting the metal-rich composition
Figure 2: XPS spectra from W 4f in a) and C 1s in b) of the
W-B-C films deposited at 500 oC and with TMB flows from bottom to the top: 1, 2.5, 5, 7.5 and 10 sccm. Literature binding energies for W-W [38] and C-W [24,39] are indicated by the vertical dotted lines in a) and b), respectively.
Intensity (arb. units)
40 38 36 34 32 30 28
Binding Energy (eV) W 4f 5p3/2 4f5/2 4f7/2 a) 10 sccm 7.5 sccm 5 sccm 2.5 sccm 1 sccm
Intensity (arb. units)
292 288 284 280
Binding Energy (eV) C 1s b) 10 sccm 7.5 sccm 5 sccm 2.5 sccm 1 sccm
6
determined for the reactively sputtered films. Measurements by Greczynski et
al. on sputtered W2C films with a
hexagonal crystal structure give a binding energy of 31.8 eV for the W4f7/2 peak [39], i.e. a chemical shift of
0.4 eV to higher binding energy compared to in our films, probably due to the higher C content in the analyzed W2C films. There is no double-peak
around 32.8-35.8 eV for W-O bonding [34]. Instead, the low intensity structure located at 36.8 eV originates from W
5p3/2 states.
The C 1s peaks in Fig. 2b are located at 283.4 eV, which is close to the reported binding energy between 283.4 and 283.6 eV reported for sputtered W2C films with a hexagonal crystal structure by Greczynski et al. [39] as well as
Liu et al. for W-B-C films [25]. The average of these measurements is indicated by a vertical dotted line at 283.5 eV in Fig. 2b. and potential C-W* bonding [40]. In addition, the C 1s spectra show no evidence of C-O bonding in the deposited films as no peak is visible at ~292 eV [34]. The B 1s peaks in Fig. 3a, centered at 187.9 eV, are of low intensity for films deposited with 1 and 2.5 sccm TMB. This binding energy is identical to the measured B-W binding energy for W2B5 [34] and W-B-C films deposited by Alishahi et al. [26], and close to the 188.1 eV
reported by Liu et al. for W-B-C films [25]. There is no indication of B-B or B-O bonding in the films as no B 1s peaks are visible at 189.4 eV [34] or ~192-193 [34], respectively.
The O 1s peaks in Fig. 3b display low intensities and asymmetric tails towards higher binding energy. The peak position located at ~530.6 eV is identical to that determined for WO3 and that
is indicated by a vertical dotted line in the spectra [34]. No corresponding peaks from oxygen bonding are observed in the W 4f, C 1s, and B 1s spectra.
The binding energies for the W 4f7/2, W 4f5/2, C 1s, B 1s, and O 1s peaks remain constant for
films deposited at RT, 100, 200, 300, 400, and 600 oC and using 10 sccm TMB in the plasma
compared to the films deposited at 500 oC with 10 sccm TMB. For the film deposited at RT
conditions this is a support of the strong driving force for W to form carbides and borides. When increasing the temperature to 700 oC, Si is detected by XPS.
3.2. Phase distribution in the deposited films
Fig. 4a shows θ/2θ diffractograms from W-B-C films deposited with TMB flows of, from bottom to the top: 1, 2.5, 5, 7.5, and 10 sccm at 500 oC. The peaks in the diffractogram
originating from the W-B-C films are broad and generally of low intensities, indicating a nanocrystalline microstructure. In addition, there are peaks from the Si(100)substrate dominated by the 400 peak at 2q ≈ 69o and the much weaker 600 peak at 2q ≈ 117o. For some
of the films the “forbidden” 200 peak at 2q ≈ 33o [41] is visible as well as an artefact from the
applied diffractometer at 2q ≈ 17o. The film deposited with 1 sccm TMB in the plasma show
peaks at the 2q angles ~40°, ~58°, ~87°, and ~101°, corresponding to the 110, 200, 220, and
Figure 3: B 1s in a) and O 1s in b) of XPS spectra from the
W-B-C films deposited at 500 oC and with TMB flows from bottom to the top: 1, 2.5, 5, 7.5 and 10 sccm. Literature binding energies for B-W [33] and O-W [33] are indicated by the vertical dotted lines in a) and b), respectively.
Intensity (arb. units)
192 188 184 180
Binding Energy (eV) B 1s a) 10 sccm 7.5 sccm 5 sccm 2.5 sccm 1 sccm
Intensity (arb. units)
538 536 534 532 530 528 526
Binding Energy (eV) O 1s b) 10 sccm 7.5 sccm 5 sccm 2.5 sccm 1 sccm
7
310 peaks from W [42] and with the 2q angles for W determined in literature indicated by dotted vertical lines in Fig. 4a. There is a peak at 2q ≈ 43° from the cubic WC1-x phase, which is the 200 peak
as calculated from the a-axis lattice parameter 4.2355 Å [43] using Cu-Kα radiation instead of Co-Kα as in
ref. [43]. The position of the 200 peak is indicated as a dotted line in the diffractogram. The presence of WC1-x in the film shows a strong
driving force towards carbide formation at low flows of TMB and is supported by the XPS analysis in section 3.1. Increasing the flow of TMB to 2.5 sccm broadens and reduces the number of peaks in the diffraction pattern, while the peaks from the W phase are retained. Increasing the TMB flow further to 5 sccm results in two visible peaks originating from WC1-x corresponding to the 200 and 400 peaks at 2q angles of ~42° and ~95°, respectively.
A strong peak at a 2q ~40° is reported in the publications by Alishahi et al. [26] and Debnárová
et al. [27] for more boron and carbon rich W-B-C films and with a second peak at 2q ~70° [26]
that is not visible for our deposited films. From the diffractogram in Fig. 4a there is no evidence of crystalline d-WC that should exhibit strong peaks at 2q ≈ 36° 1010 and 2q ≈ 48° 1011 [44], or crystalline tungsten borides such as WB2 with AlB2 type-structure (Strukturbericht
designation C32) with strong peaks at 2q ≈ 29° 0001 and 2q ≈ 46° 1011 [45], or hexagonal W2B5 with strong peaks at 2q≈25° 0004 and 2q ≈ 35 1010 and 1011 [46]. For tetragonal WB
there is an overlap between two strong peaks,105 and 112, with the WC1-x 200 peak, but with
no matching peaks expected at 2q ≈ 94°, and where a strong 103 peak at 2q ≈ 33° is absent [47]. To conclude, the θ/2θ diffraction shows no evidence for either crystalline d-WC or tungsten borides, and where the WC1-x phase is the preferred crystalline phase in agreement with what
has previously been reported for sputtered WC films [11] [14] [17] [18] [21] [23]. The relatively strong intensities of the 200 and 400 peaks in the diffractogram suggests that the WC1-x film has
corresponding preferred crystallographic orientation i.e., 100-oriented.
To investigate the film orientation with respect to the substrate, we applied pole figure measurements, see Fig. 5. As observed, the monitored {200} pole displays a point of high intensity centered in the middle of the figure where y » 0°, i.e. in the growth direction. The investigated {220} and {222} pole figures have high intensity rings located at y » 45° and y » 54.7°, respectively, which is expected for a 100-oriented fiber-textured film [48].
Additional pole figure measurements at the expected positions of {200}, {211}, {220} and, {310} in a carbon supersaturated bcc-structured a-W phase (not shown), as observed by e.g. Pauleau for W-C coatings containing less than 25 at.% C refs. [49][51], was conducted to eliminate the possibility of such a phase being present in our films.
Figure 4: (a) X-ray q/2q diffractograms recorded from W-B-C films deposited at 500 oC and with TMB flows from bottom to the top: 1, 2.5, 5, 7.5 and 10 sccm. (b) X-ray q/2q diffractograms recorded from W-B-C films deposited from bottom to the top at: RT, 100, 300, 400, 600, and 700 oC with 10 sccm TMB in the plasma.
Log intensity (arb.units)
140 120 100 80 60 40 20 Scattering Angle 2θ ( o ) Si 600 Si 400 W 110 Si 200 W 200 W 220 W 310 WC 1-x 200 WC 1-x 400 a) 500 oC 10 sccm 7.5 sccm 5 sccm 2.5 sccm 1 sccm
Log intensity (arb.units)
140 120 100 80 60 40 20 Scattering Angle 2θ ( o ) Si 200 Si 600 W 110 W 200 W 220 W 310 Si 400 WC 1-x 200 WC 1-x 400 b) 700 o C 600 o C 400 oC 300 o C 100 o C RT 10 sccm
8
At TMB flows of 7.5 and 10 sccm the 200 and 400 peaks increase in intensities and there are no indications of crystalline borides in the films even though the films contain 6-7 at.% B. This suggests that W has a stronger driving force to form carbides compared to borides in reactive sputtering with TMB, but where a solid solution of B in the WC1-x
phase cannot be excluded. This is supported by XPS in Fig. 3a that shows W-B bonding.
Fig. 4b shows θ/2θ diffractograms from films deposited at deposited at from bottom to the top: RT, 100, 300, 400, 600, and 700 oC and with 10 sccm TMB in the plasma. Already for growth
at RT conditions there is a broad peak centered at 2q≈38o beside the peaks previously described
from the Si(100) substrate in Fig. 4a. The peak is centered at a lower diffraction angle compared to the literature values for the W 110 peak at 2q ≈ 40.26o [42] and the WC
1-x 200 peak at 2q ≈
42.66o, i.e. at the right-hand side of the peak and with the literature values determined for W
and WC1-x indicated as dotted line in the diffractogram. The peak position is close to
orthorhombic [51] and hexagonal [52] W2C at with peaks
2q ≈ 38o, but where the absence of other peaks makes phase
identification uncertain. In addition, we note the possibility of a supersaturated solid solution of C in a-W (bcc) [21][49][50], and/or an amorphous phase content.
The peak remains centered at 2q ≈ 38o when the substrate
temperature is increased to 100 oC and 200 oC (not shown).
For additional investigation of the phase distribution in these films, we applied diffraction in GI geometry.
Figure 6 shows GI-XRD measurement of a W-B-C film deposited at 200 °C and with a 10 sccm TMB flow, revealing two broad structures centered around 2q ≈ 38o and
2q ≈ 72o. These angles are close to expected positions for
high intensity peaks of both orthorhombic [51] and
hexagonal [52] W2C. The corresponding q/2q
diffractogram in Fig. 6 provides no help for determining the W2C polytype as the substrate 400 diffraction peak overlap with the peaks at 2q ≈ 72o. For the
broad peak at 2q ≈ 38o, both geometries suggest a phase mixture with W2C, possible
supersaturated solid solution of C in a-W (bcc) or WC1-x contributions for films deposited at
100 oC (not shown) and 200 oC with 10 sccm TMB in the plasma.
At a deposition temperature of 300 oC the peak at 2q ≈ 38o in Fig. 4b shifts to a higher diffraction
angle of ~42°, and a second peak is found at ~92°. These are the 200 and 400 peaks from WC 1-x and indicate an oriented growth to the Si(100) substrate already at 300 oC. Increasing the
growth temperature to 400, 500 (in Fig. 4a) and 600 oC results in increasing intensities for the
WC1-x 200 and 400 peaks. However, at 700 oC the diffraction pattern completely changes seen
from the number of peaks and their positions. This is attributed to a reaction with the substrate, where Si is diffusing into the film and crystalline tetragonal WSi2 [53] is formed. The WSi2
Figure 5: Pole figure measurements of the 200, 220, and 222
reflections for a W-B-C film deposited at 500 °C and with a 10 sccm TMB flow showing a 100-oriented fiber texture.
Figure 6: q/2q and grazing incidence (GI) (q=2°) XRD measurements of a W-B-C film deposited at 200 °C and with a 10 sccm TMB flow. The q/2q diffractogram is vertically shifted for clarity. Intensity (log) 120 100 80 60 40 20 Scattering Angle 2θ ( o ) GI θ/2θ Si 600 Si 400 W2 C W C1-x 200 W 110 W2 C W 211 TMB 10 sccm at 200 oC
9
peaks are seen from increasing 2q: 002, 101, 110, 103, 112, 200, and 202/114 at 2q ≈ 63o, and
where the intensity distribution for the peaks indicate a random orientation for the film. Silicide formation during sputtering of W on Si(100) substrates has previously been reported following annealing at 1073 K (800 oC) [54]. The diffraction peaks from WSi
2 gain in intensities at 800 oC, but reduces in intensities at 900 oC to become extinct, i.e. an X-ray amorphous film was
formed (not shown). For this film, XPS measurements show a surface composition with a Si content of ~96 at.%.
3.3. Morphology of the films
Figure 7 shows SEM images obtained from films deposited at TMB flows of 1, 2.5, 5, and 10 sccm at 500oC in (a) and films deposited at 200, 400, 600, and 900 oC and with 10 sccm TMB
in the plasma in (b). The films deposited with different TMB content in the plasma exhibit thicknesses in the range ~800 to ~1000 nm corresponding to a deposition rate of 160 to 200 nm/min and with higher thicknesses for the films deposited with 5 and 10 sccm TMB in the plasma. The microstructure changes from broken columns, as in the film deposited with 1 sccm TMB in the plasma to fine-grained in the film grown with 10 sccm. A fine-grained microstructure for WC1-x films finds support from literature, where WC1-x films are often
described as being nanocrystalline/ fine-grained [14] [23]. In addition, Mitterer et al. showed that alloying TiC with B favor a fine-grained microstructure [55]. The surface structure of the films is generally smooth, which is supported from ocular inspection of films that are the silver-white with a metallic luster. Furthermore, the WC1-x
films deposited at 400 and 600 oC in Fig. 7 also
exhibit a fine-grained microstructure c.f. Fig.
4b. Decreasing the
deposition temperature to 200 oC results in a more glass-like microstructure possibly associated with the growth of W2C,
see Fig. 4b and Fig. 7, but where the film maintain its silver-white color and metallic luster. Increasing the deposition temperature to 900 oC results in a fine-grained microstructure and
with an increased variation of the thickness in the film as well as increased surface roughness. This is accompanied by a change to a greyish colored surface. Our XRD analysis in section 3.2 explained this being the result of a reaction between the film and the Si(100) substrate. Such reaction between the film and the substrate is likely to compromise the properties of the films such as the H and Er.
3.4 Mechanical properties
The H and Er of the films were assessed by nanoindentation with films deposited at TMB flows
of 1, 2.5, 5, 7.5, and 10 sccm at 500 oC in Fig. 8a and films deposited at 200, 300, 400, 500,
600, 700, 800, and 900 oC and with 10 sccm TMB in the plasma in Fig. 8b. The measured H
and Er values show no clear trends for the films deposited at different TMB flows in the plasma.
In contrast, both the H and the Er decrease markedly by a factor of three to four for the films
deposited at 700, 800, 900 oC that is accompanied by an increased spread between the individual
Figure 7: (a) SEM images obtained from W-B-C films deposited at 500oC and with TMB flows of 1, 2.5, 5, and 10 sccm. (b) SEM images obtained from W-B-C films deposited at 200, 400, 600, and 900 oC with 10 sccm TMB in the plasma.
200 nm 200 nm 200 nm 200 nm TMB 1 sccm TMB 2.5 sccm TMB 5 sccm TMB 10 sccm 200 oC 900 oC 400 oC 600 oC 200 nm 200 nm 200 nm 200 nm
10
measurements during nanoindentation. This is a consequence of the reaction between the Si(100) substrate and the film, resulting in formation of WSi2 at 700 and 800 oC or a Si rich
surface at 900 oC as determined from our XRD and XPS measurements and where the reaction
also yields an increased surface roughness seen from the loss of the silver-white color and metallic luster. The H varies in the range 23 to 31 GPa for films deposited with different TMB flows and at temperatures ≤ 600 oC; a closer inspection shows the lowest H values for the films
deposited at temperatures of 200, 300 and 400 oC, with H = 23, 25, and 24 GPa, respectively.
Our H values are comparable or even slightly higher than the reported HV of d-WC with 22 GPa for a 0001-oriented single-crystal [2] as well as to those reported for more carbon and boron rich W-B-C films by Alishahi et al. [26] and Debnárová et al. [27] with values in the range ~24 to ~29 GPa and ~23 GPa, respectively. In contrast, our H values are lower than the ~37 to ~47 GPa reported by Su et al. [24] for d-WC films alloyed with ~4 to ~7 at. N as well as the ~45 GPa determined by Liu et al. for W-B-C films reactively sputtered from a WB2 target
[25]. The H values measured by Su et al. support the superior mechanical properties of d-WC films compared to WC1-x films in particularly when d-WC is alloyed with N. For the W-B-C
films reported by Liu et al. it is important to firstly note the higher boron content in their films with ~50 at% compared to our films with 6-8 at.% B. Secondly, the phase distribution in their films is different with peak(s) from crystalline WB2 with C32 crystal structure [25], while
diffractograms recorded from our films show crystalline WC1-x with a possible solid solution
of B. This suggests less favorable hardness in WC1-x
films compared to WB2
films as H values of ~40 GPa have been measured for monolithic WB2 films
[45], but where bulk W2B5
is reported to exhibit a HV value of 26.1 GPa [2], i.e. comparable to our films. Furthermore, the measured H values are higher than the ~17 GPa obtained by
Palmquist et al. for
polycrystalline W2C and WC1-x films with ~22 at.% C [14]. From this study we note the higher
H value of 34.5 GPa measured for an epitaxial W film with 7 at.% C, suggesting that the H depends on film orientation. Indentation, including HVmeasurements, of more carbon rich films show scattered values of ~17 GPa [23] and 3500 to 4500 HV0.025 [17] corresponding to
~32 to 41 GPa using the correction factors described in the Experimental details section, but where the authors acknowledge that the measured H values are likely affected by the low thickness of the investigated films [17]. For sputtered WC/a-C:H films, Drábik et al. presents H values in the region ~13 to 15 GPa [20], showing that growth of nanocomposites results in reduced hardness in the W-C-H system. Thus, the measurements show that our reactively sputtered W-rich films exhibit comparable or even superior H values to bulk d-WC as well as to more carbon and/or boron rich films.
The Er measured in our films follow the same trends previously described for hardness, see Fig.
8a and 8b, with the highest value of ~ 280 GPa for the film deposited with 10 sccm TMB at 500
oC and a lowest value of ~130 GPa for the film deposited at 900 oC. Similar as for the hardness,
Figure 8: Mechanical properties, H as darker dashed lines and Er lighter dotted lines, for W-B-C films deposited at 500oC and with TMB flows of 1, 2.5, 5, 7.5, and 10 sccm in (a) and for films deposited at 200, 300, 400, 500, 600, 800, and 900 oC and with 10 sccm TMB in the plasma in (b). The lines are guides for the eye.
40 30 20 10 0 H (GPa) 12 10 8 6 4 2 0 TMB flow (sccm) 400 300 200 100 0 Er (GPa) T=500 oC a) 40 30 20 10 0 H (GPa) 1000 800 600 400 200 Temperature (oC) 400 300 200 100 0 Er (GPa) TMB 10 sccm b)
11
our Er ~ 280 GPa value is similar to the Er ~ 300 GPa reported by Debnárová et al. [27] but
lower than the ~ 450 GPa reported in the studies by Alishahi et al. [26]. Furthermore, for W2C
and WC1-x films containing 35 at.% C Palmquist et al. [14] measured an E-moduli of ~ 450
GPa, i.e. similar to that of Alishahi et al. We attribute the lower Er values of our films compared
to Alishahi et al. and Palmquist et al to the lower carbon and boron contents in our films with ~ 18 at.% and ~ 6 at.%, respectively.
4. Conclusions
Reactive sputtering of W in Kr/TMB plasmas results in growth of W-rich 100-oriented WC1-x
with a potential boron solid solution. The applied TMB flow with ~93 at.% W at 1 sccm and ~72 at.% W at 10 sccm determines the metal content for films deposited ≤ 600 oC. The C and
B contents in such films reach a maximum of ~18 at.% and ~7 at.% respectively and where both elements are chemically bonded to W. Growth conditions with TMB flows ≥ 5 sccm and temperatures ≤ 600 oC result in deposition of crystalline and 100-oriented WC
1-x with a possible
solid solution of Bwhereas lower TMB flows give W films and temperatures below 300 oC
yield W2C films. For films deposited at different TMB flows and temperatures ≤ 600 oC,
nanoindentation shows hardness values in the range ~23 to ~31 GPa and reduced elastic moduli between ~220 and 280 GPa. Temperatures above 600 oC leads to the formation of WSi
2 as
introduced by a reaction between the film and the Si(100) substrate. This reaction changes the color of the films from silver-white with metallic luster to greyish and with increased surface roughness as well as reduces the H and Er of the films by a factor of three to four.
Acknowledgements
The research leading to these results has received funding from the Swedish Government Strategic Research Area in Materials Science on Advanced Functional Materials at Linköping University (Faculty Grant SFO-Mat-LiU No. 2009-00971). MM acknowledges financial support from the Swedish Energy Research (no. 43606-1), VINNOVA ( 04410, 2018-04417), the Swedish Foundation for Strategic Research (SSF) (no. RMA11-0029) through the synergy grant FUNCASE and the Carl Tryggers Foundation (CTS16:303, CTS14:310). GG thanks the Knut and Alice Wallenberg Foundation Scholar Grant KAW2016.0358, the VINN Excellence Center Functional Nanoscale Materials (FunMat-2) Grant 2016-05156, and the Åforsk Foundation Grant 16-359. Harri Savimäki is acknowledged for construction of the gas handling system.
References
[1] T. Massalski, J. Murray, L. Bennett and H. Baker, Binary alloy phase diagrams, Metals Park, OH: American Society for Metals, 1986.
[2] H. O. Pierson, Handbook of refractory metal carbides and nitrides: properties, characteristics, processing and applications, NJ: Noyes publications, 1996.
[3] N. Lundberg, M. Östling, C.-M. Zetterling, P. Tägtström and U. Jansson, CVD-Based Tungsten Carbide Schottky Contacts to 6H-SiC for Very High-Temperature Operation, J. Elec. Mat., 29 (2000) 372-375.
[4] T. Takahashi and H. Itoh, Chemical vapor deposition of tungsten carbide dendrites, J. Cryst. Growth, 12 (1972) 265-271.
12
[5] M. Katoh and H. Kawarada, Heteroepitaxial growth of tungsten carbide films on W(110) by plasma-enhanced chemical vapor deposition, Jpn. J. Appl. Phys. 34 (1995) 3628-3630. [6] M. Fitzsimmons and V. K. Sarin, Comparion of WCl6-CH4-H2 and WF6-CH4-H2
systems for growth of WC coatings, Surf. Coat. Technol., 76-77 (1995) 250-255.
[7] P. Tägtström, H. Högberg, U. Jansson and J.-O. Carlsson, Low pressure CVD of tungsten carbides, Journal de Physique IV, Colloque C5, supplement au Journal de Physique II, C5 (1995) 967-974.
[8] H. Högberg, P. Tägtström, J. Lu and U. Jansson, Chemical vapour deposition of tungsten carbides on tatalum and nickel substrates, Thin Solid Films, 272 (1996) 116-123.
[9] W. S. Williams, Physics of transition metal carbides, Materials Science and Engineering A, 105-106 (1988) 1-10.
[10] A. K. Taylor, High speed friction and wear characteristics of vapor deposited coatings of titanium carbide and tungsten carbide plus cobalt in a vacuum environment, Thin Solid Films 40 (1977) 189-200.
[11] K. Fuchs, P. Rödhammer, E. Bertel, F. P. Netzer and E. Gornik, Reactive and non-reactive high rate sputter deposition of tungsten carbides, Thin Solid Films, 151 (1987) 383-395. [12] E. C. Weigert, M. P. Humbert, Z. J. Mellinger, Q. Ren, T. P. Beebe Jr., L. Bao and J. G.
Chen, Physical vapor deposition synthesis of tungsten carbide monocarbide (WC) thin films on different carbon substrates, J. Vac. Sci. Technol. A, 26, (2008) 23-28.
[13] K. A. Taylor, Ancillary properties of vapor-deposited carbide coatings, thin solid films, 40 (1977) 201-209.
[14] J.-P. Palmquist, Z. Czigany, M. Odén, J. Neidhardt, L. Hultman and U. Jansson, Magnetron sputtered W-C films with C60 as carbon source, Thin Solid Films 444 (2003) 29-37.
[15] J.-P. Palmquist, Z. Czigány, L. Hultman and U. Jansson, Epitaxial growth o tungsten carbide films using C60 as carbon precursor, J. Cryst. Growth, 259 (2003) 12-17.
[16] P. K. Srivastava, V. D. Vankar and K. L. Chopra, High rate reactive magnetron sputtered tungsten carbide films, J. Vac. Sci. Technol. A, 3 (1985) 2129-2134.
[17] G. Keller, R. Erz, I. Barzen, M. Weiler, K. Jung and H. Ehrhardt, Mechanical properties, structure and composition of ion-plated tungsten carbide films, Vacuum, 41 (1990) 1294-1296.
[18] G. Keller, I. Barzen, R. Erz, W. Dötter, S. Ulrich, K. Jung and H. Ehrhardt, Crystal structure, morphology and composition of magnetron sputtered tungsten carbide films, Fresenius J. Anal. Chem. 341 (1991) 349-351.
[19] C. Rebholz, J. M. Schneider, A. Leyland and A. Matthews, Wear behavior of carbon-containing tungsten coatings prepared by reactive magnetron sputtering, Surf. Coat. Technol., 112 (1999) 85-90.
[20] M. Drábik, M. Truchlý, V. Ballo, T. Roch, L. Kvetková and P. Kúš, Influence of substrate material and its plasma pretreatment on adhesion and properties of WC/a-C:H nanocomposite coatings deposited at low temperatures, Surf. Coat. Technol. 333 (2018) 138-147.
[21] P. Gouy-Pallier and Y. Pauleau, Tungsten and tungsten-carbon thin films deposited by magnetron sputtering, J. Vac. Sci. Technol. A, 11 (1993) 96-102.
[22] P. D. Rack, J. J. Peterson, J. Li, A. C. Geiculescu and H. J. Rack, Thin film growth of reactive sputter deposited tungsten-carbon thin films, J. Vac. Sci. Technol. A 19 (2001) 62-65.
[23] N. Radić, B. Gržeta, O. Milat, J. Ivkov and M. Stubičar, Tungsten-carbon films prepared by reactive sputtering from argon-benzene discharges, Thin Solid Films 320 (1998) 192-197.
13
[24] Y. D. Su, C. Q. Hu, C. Wang, M. Wen and W. T. Zheng, Realatively low temperature synthesis of hexagonal tungsten carbide films by N doping and its effect on the preferred orientation, phase transition and mechanical properties, J. Vac. Sci. Technol. A, 27 (2009) 167-173.
[25] Y. M. Liu, Z. L. Pei, J. Gong and C. Sun, Effect of carbon content on microstructures, mechanical properties and tribological properties and thermal stability in WBC films, Surf. Coat. Technol., 291 (2016) 276-285.
[26] M. Alishahia, S. Mirzaei, P. Souček, L. Zábranský, V. Buršíková, M. Stupavská, V. Peřina, K. Balázsi, Z. Czigány and P. Vašina, Evolution of structure and mechanical properties of hard yet fracture resistant W-B-C coatings with varying C/W ratio, Surf. Coat. Technol. 340 (2018) 103–111.
[27] S. Debnárová, P. Souček, P. Vašina, L. Zábranský, V. Buršíková, S. Mirzaei and Y. T. Pei, The tribological properties of short range ordered W-B-C protective coatings prepared by pulsed magnetron sputtering, Surf. Coat. Technol. 357 (2019) 364–371. [28] J. S. Lewis, S. Vaidyaraman, W. J. Lackey, P. K. Agrawal, G. B. Freeman and E. K.
Barefield, Chemical vapor deposition of boron-carbon films using organometallic reagents, Mat. Lett. 27 (1996) 327-332.
[29] M. Chubarov, H. Pedersen, H. Högberg, V. Darakchieva, J. Jensen, P. O. Å. Persson and A. Henry, Epitaxial CVD growth of sp2-hybridized boron nitride using aluminium nitride as buffer layer, Phys. Stat. Solidi RRL 5 (2011) 397-399.
[30] M. Chubarov, H. Pedersen, H. Högberg, J. Jensen and A. Henry, Growth of high quality epitaxial rhombohedral boron nitride, Cryst. Growth Des. 12 (2012) 3215-3220.
[31] M. Chubarov, H. Pedersen, H. Högberg, Z. Czigany and A. Henry, Chemical vapour deposition of epitaxial rhombohedral BN thin films on SiC substrates, Cryst. Eng. Comm. 16 (2014) 5430-5436.
[32] H. Högberg, J. Birch, M. P. Johansson and L. Hultman, Deposition of epitaxial transition metal carbide films and superlattices by simultaneous direct current metal magnetron sputtering and C60 evaporation; J. Mater Res. 16, (2001) 633-643.
[33] S. Hüfner, Photoelectron spectroscopy: Principles and applications, 3rd edn., Berlin, Germany: Springer-Verlag, (2010).
[34] J. F. Moulder, W. F. Stickle, P. E. Sobol and K. D. Bomben, Handbook of X-ray photoelectron spectroscopy. Perkin-Elmer Corporation, Eden Prairie, USA, (1992). [35] W. C. Oliver and G. M. Pharr, W. Oliver and G. Pharr, Measurement of hardness and
elastic modulus by instrumented indentation: Advances in understanding and refinnements to methodology, J. Mater. Res. 19 (2004) 3-20.
[36] E. Broitman, Indentation hardness measurements at macro-, micro-, and nanoscale: a critical overview, Trib. Lett. 65, (2017) 23.
[37] ISO 14577-4:2015, Metallic materials - Instrumented indentation test for hardness and materials parameters, International Organization for Standardization, Geneva, Switzerland (2015).
[38] Thompson, A. C.; Vaughan, D., X-ray Data Booklet, Lawrence Berkeley National Laboratory University of California Berkeley, CA 94720, (2001). http://xdb.lbl.gov/. [39] G. Greczynski, D. Primetzhofer and L. Hultman, Reference binding energies of transition
metal carbides by core-level x-ray photoelectron spectroscopy free from Ar+ etching artefacts, Appl. Surf. Sci. 436 (2018) 102–110.
[40] M. Magnuson, E. Lewin, L. Hultman and U. Jansson, Electronic structure and chemical bonding of nanocrystalline-TiC/amorphous-C nanocomposites, Phys. Rev. B 80 (2009) 235108.
[41] P. Zaumseil, High-resolution characterization of the forbidden Si 200 and Si 222 reflections, J. Appl. Cryst. 48 (2015) 528-532.
14
[42] JCPDS - International Centre for Diffraction Data of W, Ref ID: 00-04-806. [43] JCPDS - International Centre for Diffraction Data WC1-x, Ref ID: 00-20-1316. [44] JCPDS - International Centre for Diffraction Data of delta-WC, Ref ID: 00-20-1047. [45] Y. M. Liu, R. Q. Han, F. Liu, Z. L. Pei and C. Sun, Sputtering gas pressure and target
power dependence on the microstructure and properties of DC-magnetron sputtered AlB2-type WB2 films, J. Alloy Comp. 703 (2017) 188-197.
[46] JCPDS - International Centre for Diffraction Data of hexagonal W2B5, Ref ID: 38-1365. [47] JCPDS - International Centre for Diffraction Data of tetragonal delta-WB, Ref ID:
35-0738.
[48] M. Birkholz, Thin Film Analysis by X-Ray Scattering, Weinheim: Wiley-VCH Verlag GmbH & Co., Germany, 2006.
[49] Y. Pauleau and Ph. Gouy-Pailler, Very hard solid-solution-type tungsten-carbon coatings deposited by reactive magnetron sputtering, Mater. Lett. 13 (1992) 157-160. [50] Y. Pauleau and Ph. Gouy-Pailler, Characterization of tungsten-carbon layers deposited on
stainless steel by reactive magnetron sputtering, J. Mater. Res. 7 (1992) 2070-2079. [51] JCPDS - International Centre for Diffraction Data orthorombic W2C, Ref ID: 20-1315. [52] JCPDS - International Centre for Diffraction Data, hexagonal W2C, Ref ID: 35-0776. [53] JCPDS - International Centre for Diffraction Data of WSi2, Ref ID: 11-0195.
[54] M. O. Aboelfotoh, Electrical characteristics of W-Si(100) Schottky barrier junction, J. Appl. Phys. 66 (1989) 262-272.
[55] C. Mitterer, P. H. Mayrhofer, M. Beschlisser, P. Losbichler, P. Warbichler, F. Hofer, N. P. Gibson, W. Gissler, H. Hruby, J. Musil and J. Vlček, Microstructure and properties of nanocomposite Ti-B-N and Ti-B-C coatings, Surf. Coat. Technol. 120-121 (1999) 405-411.
